Granular customer data will revolutionize megawatt markets.
Tim Porter (firstname.lastname@example.org) is a senior executive in Accenture’s utilities industry group. Andre Begosso (email@example.com) is a senior manager in Accenture’s utilities industry group’s management consulting practice. The authors acknowledge the contributions of Curtis Bech, Accenture consultant.
Wholesale power markets have been in flux over the past year due to many market forces and new regulations. First, the slowdown of economic activity throughout the country has reduced native loads for utilities, making more generation available to the wholesale market. Second, contracting activity in the industrial segment has picked up as some companies are looking to lock in long-term power supply contracts with local utilities and power suppliers. This is counter to past observations where industrial customers historically have been wary of long-term contracts—with an average contract churn of 17 months or less. Third, regulatory activity in many states has introduced aggressive mandates for energy efficiency (EE) and demand response (DR) to curb demand growth and reshape the load curve. Fourth, changes enacted by the Federal Energy Regulatory Commission (FERC)—like the elimination of the interruptible load for reliability (ILR) provision in PJM—brought additional DR resources to the market. Last, an increasing amount of variable generation is being connected to the bulk power system and some operational problems have begun to arise.
Adding to these changes, the deployment of smart technologies—especially smart meters—will dramatically change the utility industry over the next five to 10 years. Retail consumers will be able to monitor their power consumption. More important, the deployment of time-of-use (TOU) tariffs will reduce the asymmetry of information between utilities and customers. For the first-time, consumers will be able to make conscious decisions about their energy-consumption patterns based on almost real-time information. Smart-meter pilots across the United States have demonstrated that the implementation of automated meter infrastructure (AMI) technology, coupled with TOU tariffs, can have a significant impact in the load profile—peak loads were reduced between 8 percent and 33 percent, while the total loads were cut by up to 15 percent.
Looking five years into the future, when many large-scale deployments of AMI technology will have been implemented, utilities will be serving significantly different load profiles than they are today. The degree of change is yet to be fully understood and it will depend mostly on the degree of adoption of variable power tariffs—such as TOU and critical-peak pricing (CPP). These tariffs—as well as the expansion of traditional utility-driven load-control programs—will ensure that the capacity and energy available in the market will consist of a combination of supply and demand resources. Adding to the uncertainty is the increasing cost effectiveness and pending favorable regulatory treatment for distributed generation (DG) technologies that could drive their increased penetration.
Overall, despite these unknowns, it’s safe to say that the wholesale power market will become much more competitive than it is today. The rationale behind this statement is that investor-owned utilities (IOU), independent power producers (IPP), load-serving entities (LSE) and other traditional players in this market will see more of their capacity available because of both the reduction of peak loads and the overall stagnation or decrease in consumption. Another consequence for wholesale markets will be the shift in value of generation assets. These assets were built years ago to serve a load with characteristics that will change significantly. The market will experience a shift in capacity utilization and this will drive a change in how generation assets are valued. First, as the load curve flattens and an increasing amount of power becomes available in the marketplace, peaking capacity probably will decrease in value as the amount of generating resources available will exceed the demand for on-peak power. Second, it’s important to note that the decrease in on-peak load isn’t the only force flattening the load curve. When facing the potential for increased penetration of DR and even plug-in hybrid electric vehicles, the magnitude of off-peak loads also will increase. The direct result of this will be the increase in value of mid-merit and baseload generation assets as their average utilization likely will increase. Finally, the changing characteristics of the market also will shift the value of wholesale products (e.g., load following, full-requirements, power blocks, etc.) and how they will be structured and priced.
In a world with large-scale implementations of smart technologies, successful players in the wholesale markets will understand that their new portfolio won’t only include generation assets but also demand assets. They will master the ability to influence demand by the application of new technologies—in EE, DR and DG—and combine them with traditional supply assets—baseload, mid-merit and peaking assets—to create products to meet their customer requirements while offering a unique proposition that includes economic, societal and regulatory values. The key to success in this new wholesale space will depend on the participant’s ability to offer truly a total integrated energy-management solution rather than a simple slate of wholesale products.
Efficiency: The Not-So-New Resource
Energy efficiency is the first of two cornerstone benefits that will be delivered by smart technologies. Yet, this is far from a new concept, as these EE programs have been around for a long time. The first time around, emphasis on EE was driven by the oil crisis of the 1970s and early 1980s. Today, the goal of these programs has been to permanently reduce overall power consumption around the clock by deploying new, more energy-efficient technologies (e.g., improved residential weatherization, compact fluorescent light bulbs and high-efficiency appliances). In fact, industrial and commercial customers have explored EE initiatives for many years now and not surprisingly, most of the easily attained efficiency gains already have been achieved. Residential customers, on the other hand, have paid little to no attention to these programs. Historical results in this market segment have been characterized by low penetration rates—in the single digits—across much of the country. Accordingly, wholesale energy markets never have paid much attention to EE. Control and monitoring of small-scale programs is costly and inefficient as it has depended on individual and lengthy energy audits. Also, utilities and power providers historically have been able to view only the aggregate results, and this granularity of data simply hasn’t been accurate or timely enough to allow these demand resources to be considered in wholesale power markets.
The deployment of AMI and the capability to monitor load and consumption of individual customers at more frequent intervals will change this old reality. It will provide utilities and power providers the ability to measure the success of EE programs on a dwelling-by-dwelling basis. By monitoring program performance on an individual basis, utilities and power providers will be able to make more rapid corrections if forecasted results aren’t being achieved. Ultimately, it will give them the ability to understand individual load shapes and better segment the customer base in a much more granular and meaningful way than the traditional residential, commercial and industrial segments.
The potential for new segmentation, by load shape and not by customer class, will allow utilities and power providers to gain powerful insights into the resources necessary to serve a specific segment. Likewise, it also will let them customize EE programs to the unique needs of each customer segment so that the programs will have higher penetration rates, be more accurate, and consequently be more successful. EE programs will be used to further reshape the load curve much like a negative generation asset. Overall, smart technologies will allow EE programs to be considered reliable demand-side resources that can be instantly monitored and quickly deployed in a matter of months, rather than years. In the short-term, these insights will change how utilities and power providers choose to serve their customers. New patterns of capacity and energy bidding will emerge, generating the need for new products, which will incorporate EE as one of the resources. In the long-term, it will impact generation-resource planning and compliance with new and stricter environmental regulations.
Similar to EE programs where utilities encourage customers to undertake such activities as installing high-efficiency equipment or weatherizing their premises, conservation programs are those that depend on customers changing their behavior to reduce the overall magnitude of their energy consumption. Anecdotally, these efforts include customers who increase the temperature on their thermostats by one degree during the summer months or those who always wash their clothes with cold instead of hot water. While the maximum potential of these programs is up for debate, the main challenge for system planners and the wholesale markets is the unpredictability of these demand-side resources. Along with EE, smart technologies will allow for a better and more granular monitoring of conservation efforts.
While regional markets in New England and PJM recently allowed for the bidding of conservation programs into capacity and energy markets, challenges could arise in the years to come when these programs reach critical mass. In order to manage these disparate residential conservation resources, they will have to be more visible to system operators. Also, new tools will have to be developed to process massive amounts of data for the purposes of forecasting and billing.
They key to success for utilities and power providers leveraging EE and conservation programs to serve the wholesale power markets will be the large-scale and even real-time gathering, processing and analysis of customer load and consumption information. Such capability doesn’t currently exist in the industry at scale and will have to be acquired or developed in the years to come if companies want to be successful in the wholesale business.
DR’s Hidden Costs
Demand response is the second of two cornerstone benefits that will be delivered by smart technologies. These DR programs are used to temporarily curtail on-peak power consumption, thereby increasing the amount of capacity available during peak periods. Usually, DR programs consist of: interruptible power participation programs where industrial and commercial customers agree to curtail their load or make their back-up emergency DG available for dispatching; and residential programs where remote controlled thermostats are used to interrupt power flow to HVAC systems, pool pumps and home appliances.
One important observation is that residential DR represents only a small share of the total DR resources available in the United States. In its 2008 report, FERC highlights this phenomenon by cataloguing DR resources in 2007 (see Figure 1).
The introduction of smart technologies likely will expand the prevalence of DR in the residential segment as utility or power-service providers will be able to determine the magnitude of customer demand that can be interrupted at the time of seasonal peak loads. As a result, utilities and power providers will be able to optimize the control of both linear and non-linear loads by interrupting the power supply to individual appliances or equipment at the customer premise. Also, smart technologies will provide for two-way communications between the utility and the customer system and will support the large-scale penetration of DR that is driven by the power provider or by the customer.
Finally, smart technologies will permit the introduction and adoption of variable power tariffs—such as TOU and CPP. Although variable tariffs are not considered a classic DR resource, they allow customers to respond to price signals by curbing and shifting consumption patterns in the same way DR resources do. The introduction of these variable tariffs will allow utilities and power providers to measure the customer’s individual elasticity to different tariff structures. By understanding their individual and aggregate elasticity, utilities and power providers will be able to shape load curves through price signals.
Another key observation is that, as the incremental amount of DR resources on the system increases, the more important it will be for utilities and power providers to have a deeper understanding of DR resource economics. On the surface, DR is not much different than peaking power plants, generally used only during peak periods. In theory, DR resources also are much cheaper to install (see Figure 2) and utilize than megawatt-hours from other sources. This is because the cost of fuel is zero and the amount of variable O&M is negligible.
With lower construction costs and lower theoretical energy costs, it would be natural for these resources to become a larger part of the generation asset. Economically, they would be a better option to raise reserve margins and improve reliability across the country. Many have seen the results of recent ISO-NE and PJM capacity markets auction as evidence of this. For example, the results of PJM’s May 2009 capacity market auction showed a dramatic reduction in capacity prices—from $110/MW-day to $16.46/MW-day in just one year, and without significant new generation capacity being added to the system. However, most of the change is a result of the change in the ILR provision by FERC, which has resulted in more DR resources being bid into the auction.
Nevertheless, such results have reinforced the growing view among the Obama administration and regulators across many states that DR resources are cheaper and more environmentally friendly. Furthermore, they appear more politically palatable, as it seems reserve margins can be improved without new generation construction and associated rate increases. The problem is that if this is correct, why have we not seen a penetration of DR resources higher than 6 percent in the United States?
The answer is simple: The theoretical energy cost doesn’t fully account for the opportunity cost that a power customer needs to be paid to not consume power. For example, if an industrial customer, like an aluminum smelter, enrolls in a DR program, the true cost of this demand resource would be the opportunity cost to refrain from using electricity to produce aluminum. The diversity of energy consumers (e.g., office buildings, HVAC units, aluminum smelters, steel mills, assembly factories, universities, etc.) makes it extremely difficult to make generic calculations for this opportunity cost. However, by looking at the bidding of these resources in PJM, we can assert that DR resources have very high energy prices. More specifically, the average energy cost for DR resources in PJM in 2008 was between 6 and 10 times higher than the price of running a peaker with a 12,000 MMBtu/kWh heat rate.
In other words, customers ask for a very high price to incent them to not consume power. This makes the economics of DR much more similar to the old super-peaking concept from the late 1990s (see Figures 3 and 4).
The fact is that, while having low capacity values due to low construction costs, DR resources have extremely high energy costs. Adding to this fact, a large amount of DR deceptively can inflate reserve margins, thereby giving a false sense of oversupply and depressing capacity prices. The result of this can be a more volatile market with lower reliability and higher costs that ultimately will be borne by ratepayers.
While the implementation of smart technologies likely will lower capacity cost, it also likely will increase the overall price of energy due to the simple increase of DR resources in the system. For the players in the wholesale space, the biggest challenge will be to properly assess, evaluate and price the DR resources, while incorporating them into their asset portfolio, much like in EE.
As the forces shaping electric consumption are changing, developing accurate demand forecasts will become ever more difficult and will require granular data gathering and analysis processes to which the industry is unaccustomed. Historically, electric and gas utilities have developed and deployed aggregate load-forecasting models in order to determine the amount of capital expenditure required to meet long-term load growth. While these analytical models do consider the impact of required maintenance and retirement of generation assets, their load-forecasting capabilities are somewhat limited. Smart technologies will substantially increase the precision, while significantly increasing the complexity, of the forecast.
Smart technologies significantly will increase the transparency of the T&D system by placing remotely controlled and monitored devices on the grid. These devices will be located at customer premises, at sub-stations, or directly on feeders. In other words, system operators will have visibility and access to the grid, the power flow in the system and finally to load and consumption data on a household-by-household basis. To take advantage of this amount of information and translate it into products in the wholesale market that will reduce the customer cost of service, utilities and power providers will have to gather, process, and analyze the information being generated by the system.
As utilities and power providers attain capabilities to crunch this vast amount of data and get better visibility into their system and customer behaviors, they will be able to incorporate demand-side assets in their wholesale portfolios and increase their ability to price these assets in the market. This negative generation—EE, conservation, and DR—will become dispatchable as regular supply assets and will become even more prominent in wholesale markets.
What is important to recognize is that the physical and workforce requirements of providing these data analytical capabilities are part of the overall system costs to tapping into any potential benefits of load-management activities. Overall, the level of granularity appropriate for accurate, yet not burdensome, forecasting programs hasn’t been defined. What is clear, however, is that getting the demand forecasts right is a critical step in incorporating demand-side assets into wholesale that doesn’t allow for a large margin for error.
Once a utility or a power provider has built the infrastructure and support staff necessary for creating accurate load forecasts that incorporate both supply and demand resources, the next critical question is: What if customers fail to act and EE, conservation, or DR program goals aren’t met?
The visibility of individual loads and the ability to aggregate in meaningful segments will allow utilities to identify quickly and correct customer behavior by changing the financial incentives they pay customers as part of these load and consumption control programs, such as EE, conservation and DR. This payment to customers determines the true cost of demand-side resources and their place on the dispatch stack. Today, these incentives are at best evaluated and adjusted on a yearly true-up, but after smart technologies, these true-ups can be as frequent as each billing cycle. The corrections naturally will try to adjust the design of these incentives to promote favorable customer behaviors and responses. These efforts will increase the magnitude of data gathering and analysis efforts necessary for a high-performing utility.
Success in the New Wholesale Market
The wholesale market will become more competitive due to loss of native load, increased generation availability, and expansion of demand-side programs. However, it will remain one of the premier avenues to increase earnings and cash flow for utilities and power providers.
Today, utilities and power providers approach the wholesale power business as a bridge between generation and retail, much like the link between supply and demand. The deployment of smart technologies will blur these lines. Accordingly, the distinction between wholesale and retail markets slowly will disappear. Utilities and power providers won’t be able to conduct their operations in a siloed manner. As demand-side assets—EE, conservation, and DR—gain scale and expand, the need to simultaneously coordinate the dispatch of electric supply and demand resources will require greater and greater visibility and real-time coordination across the electricity value chain. A new operating model and business processes that will provide for better coordination of supply and demand will emerge. This new model will help companies crunch data that is generated by smart technologies, ascertain patterns in increasingly unpredictable customer energy-consumption behavior, and coordinate appropriate responses. These capabilities or skills will distinguish the winners from the rest of the pack. It’s unavoidable that over the next decade, large central-plant generation facilities will remain a dominant part of the electric-supply equation, but demand-side assets will become a greater part of the whole equation. Regardless of whether these assets are fossil-fuel, nuclear, wind-powered, EE or DR resources, the winners in a smart technologies world always will be able to optimize the entire wholesale portfolio, having a mixture of energy management and energy sales. To achieve such vision, high-performing companies will embark on a three-phase journey:
• Phase I, Build the Foundation: Change operating model and organizational structure, and implement new processes to allow the incorporation of smart technologies;
• Phase II, Grow Integrated Wholesale Portfolio: Build integrated wholesale products and services portfolio enabling deployment of supply-and-demand assets, including energy efficiency and DR resources;
• Phase III, Expand Business to Fulfill Vision: Optimize the entire wholesale portfolio, using a mixture of energy management and kilowatt-hour sales
Although this won’t happen in the next five years, smart technologies eventually will create markets comprised of one customer. Utilities and power providers will no longer segment their customer base into just a few categories. They instead will view their constituents as a large portfolio of individuals and will have access to the information necessary to tailor products and services to each of these individual customers or groups of customers. This is the future that utilities and power providers must begin preparing for today.